近红外分析中的基线漂移及校正方法

Baseline drift and calibration methods in NIR analysis

基线漂移是光谱测定中极易出现的情况,基线校正是光谱数据处理中不可缺少的关键环节。以乏燃料后处理工艺过程中的2BX料液为对象,研究了近红外光谱基线漂移对料液中各组分分析结果的影响,采用一阶微分、建立混合模型及将两者结合使用的方法对光谱数据进行预处理,建立校正模型,对比了3种方法校正后样品不同时间测定结果的RSD值。结果表明,将一阶微分和建立混合模型结合使用的方法校正效果最佳,尤其对于低浓度样品效果最为显著,校正后测定结果精度大幅提高,模型稳健性增强。

Baseline drift is common in spectroscopic analysis. It is much necessary to calibrate the baseline drift during the data preprocessing. In the present work,effects of baseline drift on the analysis of 2BX flow were studied. Methods of derivative,founding mix models and the combination of the two models were applied to calibrate the baseline drift. According to the comparision experiments and the results,the method of combination of derivative and founding mix models could get the best precision,especially for the low concentration samples,and the analytic accuracy as well as robustness of the model were improved remarkably.